We conducted a prospective observational study on the medical management and health service resource utilization associated with the hospital care of patients with community-acquired lower respiratory tract infection. Between January 1994 and June 1995, 28 such patients developed Clostridium difficile-associated diarrhoea; these 28 patients were matched with 56 age-matched patients, who were used as a control group in a comparative study. Progress during the first week after admission was similar as measured by fever days and pathology or radiology use. The use of iv cephalosporins (g/day) during the first week was greater in the group who developed C. difficile-associated diarrhoea than in controls. The length of hospital stay was 36.4 +/- 21.6 days in patients with C. difficile-associated diarrhoea compared with 19.8 +/- 13.3 days in controls. Cases also required more pathological and radiological tests and greater use of antimicrobials and other drugs; however, if pathology and radiology use was calculated per day of patient stay there was no difference between the two groups. When antimicrobial use was compared, controlling for the time taken until found to be C. difficile toxin positive, patients with C. difficile infection received more iv cefuroxime as well as more total cephalosporins, beta-lactams and macrolides measured in g/day. Interestingly, in this study we could not show an increased mortality associated with C. difficile diarrhoea despite obvious evidence of morbidity. The development of C. difficile-associated diarrhoea substantially increases health care resource utilization for individual patients who are admitted to hospital with lower respiratory tract infection.